CustomGPT, AI assistants or co-pilots… While the names of the AI assistants differ depending on the provider, the basic idea is always similar: users can create their own AI assistants for specific applications and use them either alone or in a team. Depending on the provider, the AI assistants can even be made available to the general public in so-called GPT stores.

Here is an example from me: This CustomGPT helps you to create guidelines for your own brand voice.

CustomGPT and AI assistants in detail

CustomGPTs are customized AI systems that are specifically tailored to certain needs, requirements or industries. In contrast to general AI models such as ChatGPT, which cover a wide range of topics and questions, CustomGPTs are trained on specific domains, data sources or task areas. As a result, they offer personalized answers and are better able to respond to context-specific questions and challenges and complete tasks for their users.

For example, a CustomGPT could be developed for a company, brand or discipline and integrate the knowledge, processes and vocabulary of that domain to provide relevant and accurate answers. CustomGPTs offer companies and professionals the opportunity to use AI-based assistance systems that are precisely tailored to their individual requirements.

---

---

What are good examples of CustomGPTs or AI assistants?

CustomGPTs or AI assistants are particularly suitable for repetitive tasks where the initial situation is always the same, but the execution can change depending on the content.

Here are some examples of employees from different departments who use CustomGPTs in their day-to-day work:

1. marketing

A marketing employee uses a CustomGPT that is trained to match the tone and style of the company blog. When the employee enters a new content idea, the CustomGPT can write a complete blog post that takes into account both the corporate language and SEO-optimized keywords. The employee can also use the GPT to generate social media posts or email newsletters that are consistent with the brand identity.

2. communication

In corporate communications, a CustomGPT trained on internal communication guidelines and preferred styles could be used to write internal emails, announcements or press releases. An employee simply enters the most important information and the GPT generates the finished text, which can be forwarded immediately. This saves the employee time and ensures that all communication materials maintain the consistent corporate language.

3. HR (Human Resources)

An HR employee can use a CustomGPT that is trained on the company’s own job advertisements and recruitment process. CustomGPT automatically creates job advertisements in various formats for online job portals and social media based on the job description and requirements. The employee simply enters the specific requirements of the position and the AI assistant ensures that the ad is both attractive and complete.

4. real estate

A real estate agent uses a CustomGPT that is trained on previous real estate exposés and market trends. He enters the details of a new property and the GPT automatically creates an appealing real estate ad that is perfectly tailored to potential buyers. The AI assistant takes into account the style and tone of previous ads and ensures that important sales arguments are emphasized. The GPT could even generate e-mail responses for interested parties.

5. medical practice

A medical assistant in a doctor’s office uses a CustomGPT that is trained on the specific processes of the practice and the most important document templates. The AI assistant supports the preparation and creation of medical documents by generating medical reports or patient information based on input.

6. NGOs and NPOs

An employee of an NGO uses a CustomGPT that is trained for the organization’s specific projects and missions. It creates customized funding applications or donation appeals based on the information entered by employees. Employees simply enter details of the project in question and the funds required, and the AI assistant generates a convincing application or appeal for donations.

7. finances

A financial analyst can use a CustomGPT that is trained on the company’s analysis methods and financial reports. The employee enters raw data or key financial figures and the AI assistant uses them to create reports that are aligned with internal standards. The CustomGPT could also automatically create market reports, analyses of trends or presentations for investors and thus relieve the analyst.

8. journalism

A journalist can use a CustomGPT that is trained on previous articles and the writing style of the medium. The journalist enters only the most important points of a story, and the AI assistant creates a complete draft of the article that matches the publication in terms of style and content. The CustomGPT can also automatically generate suggestions for titles, captions or social media teasers.

9. e-commerce

An e-commerce employee uses a CustomGPT that is trained on product descriptions and customer feedback. As soon as new products are added to the store, the employee enters the basic information and the CustomGPT creates detailed and appealing product descriptions.

10. law

A lawyer uses a CustomGPT that is trained on previous contracts and pleadings. The attorney enters only the specific contract details, and the CustomGPT creates a draft contract, letter or legal document that meets the office’s current legal standards. In addition, the GPT could answer routine queries from clients in advance and support the lawyer with well-structured document templates.

---

Good news first: CustomGPTs can also be used for sensitive data

Many readers will immediately think of OpenAI when they think of CustomGPTs and then of data protection. It’s true, CustomGPTs originally came from OpenAI and the AI assistants with the GPT model are probably still the most widely used. But there are now also safe GPT alternatives from Switzerland and Germany. With these solutions, data protection is 100% guaranteed and applications with sensitive data can also be implemented with CustomGPT.

How do I create a CustomGPT or my own AI assistant?

To create a CustomGPT, you usually proceed in several steps. The process may be technically different depending on the platform or framework you are using, but here is a general guide:

1. define goal

First of all, it must be clearly defined what the CustomGPT is to be used for. You determine which specific tasks the AI should take on, e.g. writing blog posts in marketing, answering customer inquiries in a doctor’s surgery or drawing up contracts in a legal department.

2. data collection and prompt

To adapt the GPT model, a very precise prompt is required and, in the best case, sample data. This data should be specific to the task area or industry in which the CustomGPT is used. Examples of data are:

• Blog posts or articles from the company for an AI marketing assistant
• Previous contracts or legal documents for an AI legal GPT

3. fine-tuning of the CustomGPT or AI assistant

In this step, the pre-trained GPT model is further trained with the collected data. You adapt the model by teaching it the company-specific content, technical language or processes. This is done in the form of prompts. For a blog-related CustomGPT, for example, the AI assistant is given previous blog posts and style guides from the company so that it strikes the right tone.

4. test CustomGPT

After the CustomGPT has been customized or created, it is tested. Employees or developers give it typical tasks and check whether the GPT’s answers meet the requirements. This checks whether the GPT responds to requests correctly and in context and whether it writes in the correct style.

5. fine-tuning and iterations

The CustomGPT will be further optimized based on the feedback from the test phase. Additional data may need to be added or adjustments made so that the AI assistant works more accurately or responds better to certain requirements.

6. integration into the workflow

As soon as the CustomGPT is working well, the AI assistant will be integrated into employees’ day-to-day work. This sometimes requires short training sessions, but is generally a minor issue as the CustomGPTs are already developed very closely with the employees.

7. monitoring and continuous improvement

Following the introduction of CustomGPT, its performance is continuously monitored to ensure that it is working efficiently. Regular updates or new training data may be required to keep the AI-Assisnsten GPT up to date and to respond to changing requirements or new content.

I have gathered a lot of experience and best practices over the last few months and will be happy to support you in implementing your CustomGPT. I offer the following services.

---

My offers for the implementation of CustomGPTs and AI assistants

---

Which providers and platforms are there for CustomGPTs?

One of the best-known platforms for implementing CustomGPTs is that of ChatGPT or OpenAI itself. A few months after the release of ChatGPT, OpenAI also made the options for individual AI assistants available.

However, there are now other providers such as Peak Privacy, SwissGPT, Google and many others that also offer the option of CustomGPTs. The providers differ not only in price, but above all in the processing of the data entered.

While OpenAI and Google store all data in America, no data is stored by providers such as Peak Privacy or SwissGPT. This means that even companies with high data protection requirements can quickly and easily create their own AI assistant.

How is data protection regulated with CustomGPTs?

CustomGPTs or AI assistants are usually hosted by third-party providers or in cloud environments. It is therefore important to check exactly where the data is stored, who has access to it and what data is stored at all before using it.

The classic CustomGPTs from OpenAI basically have the same data protection guidelines as the general use of ChatGPT. However, there are also secure solutions, such as those from Peak Privacy or SwissGPT. These providers can also be used by companies with high data protection requirements.

You can also read more about data protection in my next article on implementing AI assistants in a data protection-safe manner.

Are there any examples of CustomGPTs?

Finally, a short video from me on the creation and use of CustomGPT.